Does it really take 21 days to break a habit?

Published 28 August 2026 · 6 min read

By Loutra · Sources (6)

The 21 days came from a surgeon, not from research

In 1960 an American plastic surgeon, Maxwell Maltz, published a book called Psycho-Cybernetics. In it he noted that his patients usually needed around three weeks to stop being startled by their own face in the mirror. He made the same rough observation about amputees and the phantom-limb sensation, and about people moving into a new house.

He was describing how long it takes to adjust to a change that has already happened — a clinical impression, in the preface of a book about self-image. Over the following decades it got flattened into "it takes 21 days to make or break a habit", which is not what he wrote and not something he tested.

The researchers who study this say so directly. Writing in the British Journal of General Practice, Gardner, Lally and Wardle put it in one sentence: "This myth appears to have originated from anecdotal evidence of patients who had received plastic surgery treatment and typically adjusted psychologically to their new appearance within 21 days." The reference attached to that sentence is Maltz's 1960 book. That is the entire evidence base.

When someone finally measured it, the answer was 66 — and 18, and 254

In 2010 Phillippa Lally and colleagues at University College London ran the study that should have existed fifty years earlier. Ninety-six volunteers each chose one everyday health behaviour — eating a piece of fruit, going for a walk, drinking a glass of water — and attached it to a single daily cue in their own life, something like "after breakfast". Then, every day for twelve weeks, they rated how automatic it felt.

Automaticity climbed along a curve: fast at first, then flattening out. The median time to reach that plateau was 66 days — about ten weeks, which is roughly what the same authors later suggested doctors should tell patients to expect.

Lally has been asked since whether her own 66 days should be treated as a rule. Her answer, in an interview with the University of Surrey, was blunt: "No. I mean for someone somewhere for one habit yes, but for most people most of the time, no."

A 2024 review of twenty studies says the same thing, louder

The obvious objection to a single 2010 study is that it is a single 2010 study. So: in December 2024, Singh and colleagues published a systematic review in Healthcare pooling 20 studies and 2,601 people, across behaviours from physical activity to flossing.

Median times to habit formation came out at 59 to 66 days. Means came out at 106 to 154 days — roughly double, which tells you there is a long tail of people for whom it takes much longer than the headline. Individual times ranged from 4 days to 335. The authors' own conclusion is not hedged: their results "clearly refute the popular notion that habits can be formed in approximately 21 days".

Missing one day does not put you back to zero

This is the part of Lally's study that almost nobody quotes, and it is the most useful thing in it. Participants who skipped a day did not lose what they had built. In the authors' words: "Missing the occasional opportunity to perform the behaviour did not seriously impair the habit formation process: automaticity gains soon resumed after one missed performance."

Be precise about what that shows and what it does not. It was measured on people learning new everyday behaviours, not on people relapsing into an addiction, and nobody has shown the two work identically. What it does dispose of is the arithmetic behind "well, that's day one again" — the idea that one miss returns you to the starting line. That never came from the research either.

Three clocks are running, and people confuse them

Most of the disappointment around day 22 comes from reading one clock and expecting another. When you stop something, three run at once.

• WithdrawalThe fastest. For nicotine, the NHS says symptoms last 3 to 4 weeks on average, though some people feel them for longer, sometimes a few months.
• AutomaticitySlower. Two months and up before the replacement behaviour runs without you managing it.
• The old associationSlowest, and it never fully clears.

That last one explains what usually feels like failure. In a 2024 review, Gardner and colleagues — Lally among them — describe unenacted habits as "weakened but not fully extinguished", still able to activate the old behaviour when the cue shows up; their example is drivers slipping back into an old route the day a road closure ends. Wanting a cigarette at a wedding two years in is not evidence that you failed. It is the cue doing what the literature says a cue does.

One caveat on all of it: none of these studies were about quitting an addiction. Lally's participants were drinking water and going for walks. Reading their timeline across to stopping smoking is an extrapolation, and we would rather say so than pretend the numbers were built for you.

What actually moves the number

The 2024 review also looked at what separated the fast habits from the slow ones, and the answers are unglamorous. Repetition mattered more than intensity — daily beats hard. Behaviours anchored to a morning routine formed more strongly than evening ones, and behaviours people chose themselves formed more strongly than ones assigned to them. So did the ones they enjoyed, which is worth remembering before picking the most punishing possible replacement.

The shape of the attempt matters too. Gardner's review separates disrupting a habit — avoiding the cue, making the thing harder to reach — from substituting one: giving the same cue a different answer. Avoidance leaves the association intact and waiting. Substitution goes at the association itself, which is why it should "minimise future lapses into unwanted habits". Your coffee is still going to arrive every morning. The question is what happens next.

None of which is a reason to count to 21, or to 66. It is a reason to stop treating a day on a calendar as a verdict.

Questions people ask

Loutra is free on the App Store

An otter that’s counting on you, a daily check-in, and a breathing round for the minutes that are hard.